Dr. Shih-Lung Shaw s Research on Space-Time GIS, Human Dynamics and Big Data

Transcription

1 Dr. Shih-Lung Shaw s Research on Space-Time GIS, Human Dynamics and Big Data for Geography Department s Faculty Research Highlight October 12, 2014 Shih-Lung Shaw, Ph.D. Alvin and Sally Beaman Professor & Arts and Sciences Excellence Professor Department of Geography and Director of Confucius Institute Center for International Education University of Tennessee, Knoxville TN [email protected] Understanding human dynamics has been a research goal pursued in many disciplines, including geography. In the past, researchers employed methods such as surveys, interviews, laboratory experiments and participatory observations to shed light on human dynamics. As the location-aware technology (e.g., Global Positioning System GPS), information and communication technology (e.g., the Internet), and mobile technology (e.g., mobile phone) became increasingly accessible to the general public in the 1990s, we began to observe major changes in how people carry out their activities and interactions as well as important impacts on the ways that economic, urban, social and political systems operate in the modern world. In the meantime, detailed data of individual activities and interactions are being collected by vendors (e.g., online purchases, retail store transactions), service providers (e.g., phone communications, credit card transactions, online social networking activities), and government agencies (e.g., smartcard for transit services, electronic toll collection, security monitoring data). These datasets cover our activities and interactions not only in physical space (e.g., physical movements) but also in virtual space (e.g., online activities and communications) at unprecedented spatial and temporal resolution levels. Deriving useful information from these big datasets to gain insight into spatiotemporal patterns of human dynamics at both disaggregate and aggregate scales presents exciting research opportunities as well as research challenges to addressing many important issues in today s world. My research interests have mainly focused on transportation, geographic information science (GIScience), and space-time analytics. With the advancements of technologies leading to important changes in human activity and interaction patterns, I began to ask the following questions: 1

2 As we become increasingly mobile and connected via modern technologies, how do human dynamics change due to activities such as telecommuting, online shopping, and online social networking? How do activities performed in virtual space (e.g., online shopping) interact with activities in physical space (e.g., shopping trips and freight shipments)? With these important changes of human dynamics, what kinds of transportation systems and services should we design and develop to meet our needs 10, 15, or 20 years down the road? Do geography and GIScience have adequate research frameworks, approaches and methods to tackle these research questions in a dynamic and space-time context, especially for managing, analyzing, modeling, and visualizing large, dynamic, spatiotemporal data at the individual level that are collected in both physical space and virtual space? Through collaborations with my students and other researchers, we have worked on a number of interesting research projects related to space-time GIS, human dynamics and big data. We have developed spatiotemporal data models, analysis functions, and visualization tools to address research questions associated with datasets ranging from individual migration history data, individual travel and activity survey data, individual GPS tracking data, taxi GPS tracking data, individual online social network data, individual mobile phone tracking data, land use data, traffic data, and satellite imageries in a space-time context. Below are some examples of our research projects: Development of a space-time GIS by extending Torsten Hägerstrand s time geography concepts to include human activities and interactions in both physical space and virtual space (see Figures 1 & 2, Shaw, S-L. and Yu, H A GIS-based time-geographic approach of studying individual activities and interactions in a hybrid physical-virtual space, Journal of Transport Geography, 17(2), and Yu, H. and Shaw, S-L Exploring potential human interactions in physical and virtual spaces: A spatiotemporal GIS approach, International Journal of Geographical Information Science, 22(4), ; also click here to access my NSF project website and download a free extension of Extended Time- Geographic Framework Tools for ArcGIS 10.) A generalized space-time path approach for studying spatiotemporal patterns of individual migration history data (see Figures 3 & 4 and Shaw, S-L., Yu, H. and Bombom, L A space-time GIS approach to exploring large individual-based spatiotemporal datasets, Transactions in GIS, 12(4), ) An Activity Pattern Analyst extension for studying individual and aggregate activity-travel patterns based on survey data collected in Beijing, China (see Figures 5 & 6 and Chen, J., Shaw, S-L., Yu, H., Lu, F., Chai, Y. and Jia, Q Exploratory data analysis of activity diary data: a space-time GIS approach, Journal of Transport Geography, 19(3), ) A spatiotemporal analytical framework for assessing the impacts of virtual communications (e.g., landline phone, mobile phone) on physical activities (see Figure 7 and Yin, L., Shaw, S- 2

3 L. and Yu, H Potential effects of ICT on face-to-face meeting opportunities: A GISbased time-geographic exploratory approach, Journal of Transport Geography, 19(3), ) Spatiotemporal analysis of critical transportation links using time geography concepts (see Figure 8 and Fang, Z., Shaw, S-L., Tu, W., Li, Q. and Li, Y Spatiotemporal analysis of critical transportation links based on time geographic concepts: a case study of critical bridges in Wuhan, China, Journal of Transport Geography, 23, ) Clustering of temporal events to reveal behavioral patterns (see Figure 9 and Pei, T., Gong, X., Shaw, S-L., Ma, T. and Zhou, C Clustering of temporal event processes, International Journal of Geographical Information Science, 27(3), ) Optimization of intra-intersection vehicle-pedestrian evacuation movements based on a space-time efficiency model (see Figure 10 and Fang, Z., Li, Q., Li, Q., Han, L. and Shaw, S-L A space-time efficiency model for optimizing intra-intersection vehicle-pedestrian evacuation movements, Transportation Research Part C: Emerging Technologies, 31, ) Estimation of reliable space-time prisms under uncertain travel time conditions (see Figure 11 and Chen, B.Y., Li, Q., Wang, D., Shaw, S-L., Lam, W.H.K., Yuan, H. and Fang, Z Reliable space-time prisms under travel time uncertainty, Annals of the Association of American Geographers, 103(6), ) Development of an object-oriented space-time raster GIS data model based on a temporal approach of organizing remote sensing datasets from multiple sources (see Figure 12 and Zhao, Z., Shaw, S-L. and Wang, D A space-time raster GIS data model for spatiotemporal analysis of vegetation responses to a freeze event, Transactions in GIS, in press, ) Use of mobile phone location data for urban land use classification (see Figure 13 and Pei, T., Sobolevsky, S., Ratti, C., Shaw, S-L., Li, T. and Zhou, C A new insight into land use classification based on aggregated mobile phone data, International Journal of Geographical Information Science, in press, ) Analysis of scio-spatio-temporal relationships among individuals in physical and virtual spaces (see Figure 14 and Yin, L. and Shaw, S-L Exploring space-time paths in physical and social closeness spaces: a space-time GIS approach, International Journal of Geographical Information Science, in press. ) Understanding of urban human dynamics using both passive mobile phone tracking data and active mobile phone tracking data (see Figures 15, 16 & 17 and click here to watch a video showing the functions available in the Space-Time Environment for Analysis of 3

4 Mobility (STEAM) for exploring and visualizing large individual mobile phone tracking data. Yang, X., Shaw, S-L., Zhao, Z., Yin, L., Fang, Z. and Li, Q Understanding of aggregate human mobility patterns using passive mobile phone location data A home-based approach, under review. and Zhao, Z., Shaw, S-L, Xu, Y. and Yin, L Cell towers as sensors: Uncovering the pulse of a city based on actively tracked mobile phone location data, under review. ) Dr. Shaw and his students/collaborators have established a Space-Time Analysis and Research on Human Dynamics (STAR-HD) research group to pursue research related to spacetime GIS, human dynamics and big data. For information of Dr. Shaw s other research areas and publications, please visit his website, access his CV, or him at [email protected]. 4

5 Figure 1: Temporal dynamic segmentation in space-time GIS for managing individual activities and interactions in both physical and virtual spaces. 5

6 Figure 2: Extended Time-Geographic Framework Tools Extension for ArcGIS. 6

7 Figure 3: Space-time paths of individual migration history data displayed in space-time GIS. 7

8 Figure 4: Generalized space-time paths showing population migration trends based on different temporal aggregation methods. 8

9 Figure 5: Space-time paths showing individual activity-travel patterns of two different neighborhoods in Beijing, China. 9

10 Figure 6: Activity density distribution of the two neighborhoods at 10 a.m. on a workday. 10

11 Figure 7: Potential face-to-face meeting opportunities between two persons with and without mobile phones. 11

12 Figure 8: Spatiotemporal pattern of weighted centers of trip origins and trip destinations using a critical bridge link during the study week. 12

13 Figure 9: Clustering results of temporal events and their dendrograms and pseudo-f statistics for discrete Fréchet (DF) distance and dynamic time warping (DTW) distance. 13

14 Figure 10: Space-time cubes at a street intersection for optimizing vehicle-pedestrian evacuation movements. 14

15 Figure 11: Reliable space-time prisms for α = 0.5 and α = 0.99, where α represents the probability of arriving at the destination before the scheduled arrival time. 15

16 Figure 12: Temporal patterns of NDVI values of 15 different vegetation classes. 16

17 Figure 13: Centers of clusters for different land-use types: (a) patterns of centers for clusters and known land-use, (b) distribution of volume for classification, (c) distribution of volume of known land use. 17

18 Figure 14: Analysis functions designed to explore space-time paths in both physical and social closeness spaces. 18

19 Figure 15: The top map shows space-time paths of selected mobile phone users in the study city. The bottom map shows the Space-Time Environment for Analysis of Mobility (STEAM) developed for users to interactively explore spatiotemporal dynamics of the entire dataset of 16 million mobile phones in Shenzhen, China. Users also could use STEAM to visualize their own datasets. 19

20 Figure 16: Density of estimated home anchor points by mobile phone tower service area. 20

21 Figure 17: Distribution of modified standard distance ( ) for 1.2 million individuals in a passive mobile phone dataset. The modified standard distance ( ) measures the spread of an individual s cell phone activities with respect to his/her home location. 21